M3−x(NH4)xCrO8 (M = Na, K, Rb, Cs): A New Family of Cr5+-Based

Feb 28, 2011 - cation, we synthesized a new class of compounds, M3-x(NH4)xCrO8 (M = Na, K, Rb, Cs). These magnetic compounds with the simple 3d1 ...
0 downloads 0 Views 2MB Size
COMMUNICATION pubs.acs.org/JACS

M3-x(NH4)xCrO8 (M = Na, K, Rb, Cs): A New Family of Cr5þ-Based Magnetic Ferroelectrics Raghabendra Samantaray,† Ronald J. Clark,† Eun S. Choi,‡ Haidong Zhou,‡ and Naresh S. Dalal*,†,‡ † ‡

Department of Chemistry and Biochemistry, Florida State University, Tallahassee, Florida 32306-4390, United States National High Magnetic Field Laboratory, Tallahassee, Florida 32310-3706, United States

bS Supporting Information ABSTRACT: Upon consideration of the hydrogen-bonding properties of the NH4þ cation, we synthesized a new class of compounds, M3-x(NH4)xCrO8 (M = Na, K, Rb, Cs). These magnetic compounds with the simple 3d1 ground state become ferroelectric. X-ray studies confirmed that the phase transition involves a symmetry change from I42m to Cmc21 to P1. The transition temperature depends linearly on the composition variable x. The transitions are of the orderdisorder type, with N-H 3 3 3 O bonding playing the central role in the mechanism. Extension of this idea to the introduction of ferroelectricity in several other classes of materials is suggested.

T

he discovery of new multiferroic materials, compounds that exhibit both magnetic and electrical ordering, is an area of high current interest in part because multiferroics offer a fundamentally new method of data storage and manipulation.1 Most of the known multiferroic materials are based on transition-metal oxides, mainly of the perovskite family ABO3, where A is a divalent metal that is usually in a diamagnetic oxidation state and B is a paramagnetic metal in a tetravalent state. The perovskite-based multiferroic field has been recently extended to compounds known as metal-organic frameworks.2 An important component of these systems is a molecule or ion that exhibits dynamic disorder at high temperatures but slows down and becomes ordered below a particular temperature Tc, leading to long-range order with electric polarization (i.e., ferroelectricity).2 In an attempt to open up a new avenue of multiferroic materials, we have investigated compounds of Cr in its (unusual) pentavalent oxidation state. These are compounds with the formula M3CrO8, (M = NH4þ, Naþ, Kþ, Rbþ, Csþ), which have been known for over a century3,4 but have only recently been reported as antiferromagnets with novel quantum critical spin dynamics.5,6 To look for ferroelectric characteristics in these materials, we first focused on (NH4)3CrO8,7 because many ionic salts of the NH4þ ion are ferroelectric, while the corresponding ones with alkali-metal ions are not.8 An example is (NH4)2SO4 versus K2SO4: the former exhibits ferroelectricity below ∼225 K, while the latter is essentially inert.8 (NH4)3CrO8 indeed exhibited ferroelectricity below 250 K, a rather elevated temperature. We then surmised that perhaps ferroelectric behavior could be induced in the other members by substituting the alkali metal with NH4þ. As shown below, this was indeed found to be the case for several new mixed lattices r 2011 American Chemical Society

M3-x(NH4)xCrO8. Variable-temperature single-crystal X-ray diffraction (XRD) studies showed that the ferroelectric transitions involve a lowering of the lattice symmetry via N-H 3 3 3 O hydrogen bonding. The transition temperature varied linearly with x, the composition ratio of the NH4þ ion, suggesting that the transition mechanism is based on the cooperativity of hydrogen bonding, providing a possible avenue for optimizing the ferroelectric properties. Preparation of the title compounds was accomplished by the reaction of H2O2 with an ice-cold solution of basic CrO3, as described previously.3,5 The mixed compounds M3-x(NH4)3CrO8 were prepared by using the appropriate molar concentrations of the alkali-metal hydroxide and NH4OH bases. Single crystals were grown by slow cooling, and their authenticity was ascertained using single-crystal XRD at various temperatures down to 100 K and powder XRD down to 6 K. Dielectric measurements were made using a standard ac bridge with an attachment for variabletemperature measurements. For the spontaneous polarization measurements, samples were poled through the phase transition temperatures at an electric field of 25 kV/cm. A Quantum Design Physical Properties Measurement System (PPMS) was used to measure the temperature dependence of the magnetic susceptibility and specific heat (Cp) at constant pressure. Figure 1a depicts the structure of the 3d1 CrO83- ion, which is the source of paramagnetism in these lattices. Figure 1b-e show schematics of the crystal structures of (NH4)3CrO8, Rb3CrO8, Rb1.36(NH4)1.64CrO8, and Cs(NH4)2CrO8, respectively. At room temperature, all of the crystals crystallize in the tetragonal space group I42m. It is known that the M3CrO8 compounds with M = Na, K, Rb, Cs (or their alloys) exhibit antiferromagnetic phase transitions only at cryogenic temperatures (below 3 K).5,6 On the other hand, in our preliminary search, (NH4)3CrO8 was found7 to exhibit a structural phase transition at 137 K, but no further attempts to search for ferroelectricity were made at that time. We thus first investigated (NH4)3CrO8 for its dielectric behavior, because as mentioned earlier, other types of NH4þ-containing lattices exhibit a multitude of dynamic phenomena and even ferroelectricity.8 Before examining the dielectric properties of (NH4)3CrO8, we measured its heat capacity, since all ferroelectric transitions exhibit peaks in heat capacity as well.8 As Figure 2a,b shows, (NH4)3CrO8 exhibits at least two peaks (i.e., phase transitions), at Tc1 = 250 K and Tc2 = 137 K, whereas Rb3CrO8 shows none. Figure 2c,d depicts the effect of substituting Rbþ and Csþ with NH4þ: Rb1.36(NH4)1.64CrO8 and Cs(NH4)2CrO8 exhibit peaks at 147 and 158 K, respectively, documenting that the replacement Received: December 31, 2010 Published: February 28, 2011 3792

dx.doi.org/10.1021/ja1117683 | J. Am. Chem. Soc. 2011, 133, 3792–3795

Journal of the American Chemical Society

COMMUNICATION

Figure 1. Structures of (a) peroxychromate, CrO83-, having a 3d1 paramagnetic core; (b) (NH4)3CrO8; (c) Rb3CrO8; (d) Rb1.36(NH4)1.64CrO8; and (e) Cs(NH4)2CrO8. H and O atoms have been omitted for clarity. In (d), NH4þ ions occupy the N2 sites, whereas the N1 sites are shared by Rbþ and NH4 þ ions in a 2:1 ratio because of their comparable sizes. In (e), however, the NH4þ ions occupy the N2 sites with double occupancy at four faces, but because of its larger size, Csþ occupies the slightly larger N1 sites on the edges and the top and bottom faces.

Figure 2. Temperature dependence of the heat capacities of the various M3-x(NH4)xCrO8 lattices.

of Mþ by NH4þ induces phase transitions. Notably, the transition temperature Tc increases in proportion to the relative amount of NH4þ to Mþ ions in the formula (see Figure 5). The clear λ-type shape of the Cp peak for (NH4)3CrO8 provided a clue to the nature of the phase transition: a typical order-disorder type. A good test for such a mechanism is that the transition entropy ΔS should be R ln n, where n is the number of sites between which the molecular entity hops in the disordered phase. Thus, ΔS over each peak was calculated by graphing Cp/T R versus T and then integrating, as ΔS = (Cp/T) dT. For the 250 K peak in (NH4)3CrO8, the value of ΔS was found to be 5 ( 0.05 J mol-1 K-1. For a two-state order-disorder mixed system, the theoretical value of ΔS is R ln 2 = 5.76 J/mol K.8 Our experimental ΔS lies within ∼15% of this value, strongly suggesting that the 250 K peak corresponds to the dynamical slowing (below 250 K) of the correlated motion of the NH4þ and CrO83- ions between two sites as a result of hydrogen-bonding effects (see below). To examine whether the transitions induced by NH4þ induce ferroelectricity, we measured the dielectric constant. Figure 3a shows the temperature dependence of ε0 for (NH4)3CrO8. A sharp λ-type peak at 250 K implying a dielectric phase transition at Tc = 250 K, together with other anomalies at lower temperatures, can be clearly seen. In this report, we focus only on the 250 K transition, since this transition shows clear signs of ferroelectric behavior. The Tc value of 250 K is in full agreement with the Cp data in Figure 2a. Definitive evidence of ferroelectricity

Figure 3. Temperature dependence of (a) dielectric constant and (b) spontaneous polarization of (NH4)3CrO8. The insets in (a) and (b) respectively, show the dielectric anomaly and the critical behavior of the pyroelectric current in the vicinity of the phase transition temperature.

(or antiferroelecity) is the observation of a pyroelectric current below the Tc of a lattice.8 Figure 3b shows the pyroelectric data: the pyroelectric coefficient γ and spontaneous polarization Ps exhibited sharp increases below 250 K when a constant heating rate was applied, as expected for a ferroelectric phase transition. A maximum polarization of 0.25 μC/cm2 was obtained, which is comparable to those of other known ferroelectrics of similar kind.8 The observation of the λ shape of the dielectric anomaly and the typical shape of the polarization curve are clear evidence of a ferroelectric transition. To examine the transition mechanism at the atomic level, detailed variable-temperature powder diffraction measurements were made down to 6 K. As shown in Figure 4a, at 250 K the 220 Bragg peak splits into two peaks (indexed as 040 and 400), indicating a tetragonal to orthorhombic phase transition at that temperature, in agreement with the transition temperature observed via Cp measurements (Figure 2a). Figure 4b shows the temperature dependence of the lattice constants in the various phases. It can be seen that there is an abrupt change in the crystal symmetry and lattice parameters at 250 K, with the space group being lowered from tetragonal to orthorhombic. The transition temperature is in conformity with the those obtained from the Cp and dielectric data. There is a further 3793

dx.doi.org/10.1021/ja1117683 |J. Am. Chem. Soc. 2011, 133, 3792–3795

Journal of the American Chemical Society

COMMUNICATION

Figure 5. Evidence of the linear increase in the ferroelectric transition temperature, Tc, of M3CrO8 as a function of [NH4þ]. Symbols are (1) Rb3CrO8, (2) (NH4)1.64Rb1.36CrO8, (3) (NH4)2CsCrO8, and (4) (NH4)3CrO8.

Figure 4. (a) XRD scans of typical peaks (as marked) of (NH4)3CrO8 at several temperatures. The 220 Bragg peak splits into 400 and 040 peaks at 250 K. Circles are experimental data, and the red and blue lines are Lorentzian fits. (b) Lattice parameters as functions of temperature.

change at 137 K, as also seen in the specific heat data. However, since the polarization current showed no change at 137 K, this phase transition is strictly structural. For further details, single-crystal X-ray studies were undertaken at the three temperatures 263, 203, and 123 K, each of which corresponds to a different phase (phases I, II, and III, respectively). The X-ray analysis clearly showed that on cooling below 250 K, the higher-temperature tetragonal phase (I42m) changed to the orthorhombic, polar phase Cmc21. The transition between phases II and III further lowered the symmetry to P1. These phase changes were found to be fully reversible with temperature. It can be seen (Figure 1) that phase I consists of two crystallographically independent NH4þ groups (containing the nitrogen atoms N1 and N2) and a CrO8 unit. N1 is located at the 4 symmetry center and N2 at 42m. There is a considerable difference between the two nitrogen atoms. There are eight N1 sites, located on the four faces (ac and bc planes) 2.811 Å away from 12 neighboring oxygen atoms. N2 sites are located on the c axis and ab plane and occupy more space. Each has eight nearest oxygen atoms at 2.964-3.012 Å. Below the phase transition, in

phase II, the ammonium group at N1 was found to have a series of N-O distances ranging from 2.796 to 3.24 Å, in contrast to phase I, wherein all these bonds were equal. Similarly, N2 shows N-O bond distances from 2.787 to 3.361 Å. Despite the fact that the X-ray study was (as usual) unable to locate exactly the NH4þ H atoms, these results do reveal that there is a significant distortion of the NH4þ ion during the phase transition due to the N-H 3 3 3 O bonding. This effect is reminiscent of the mechanism of the wellestablished antiferroelectric phase transition in NH4H2PO4.8,9 Similar X-ray diffraction, heat capacity, and dielectric measurements showed that the mixed lattice of Rb1.36(NH4)1.64CrO8 exhibits ferroelectricity below 147 K, while Cs(NH4)2CrO8 becomes ferroelectric below 158 K (see Figure 2). An important observation was that the plot of Tc versus x in the above-discussed M3-x(NH4)xCrO8 lattices is strictly linear, as shown in Figure 5, suggesting that the cooperativity leading to the ferroelectric transitions in these materials indeed arises through the N-H 3 3 3 O hydrogen-bonding interaction. It would thus be of future interest to examine the effect of H f D isotope substitution on Tc and the saturation polarization Ps. The results suggest a possible avenue for creating more desirable characteristics. Such isotope effects are well-known in other ferroelectrics, such as the KH2PO4 family.8 In summary, we have synthesized a new class of magnetic ferroelectrics through a property-by-design approach, inducing phase transitions and dielectric polarization via hydrogen bonding. This procedure should be effective for generating ferroic characteristics in other families, such as the alkali-metal peroxyniobates (M3NbO8) and peroxytantalates (M3TaO8).4 It certainly broadens the scope of the chemistry of Cr in its unusual oxidation states. The idea ought to apply also to the MCuCl3 family (M = alkali metal), which is known to exhibit very unusual transitions (e.g, with quantum steps in magnetization),10 but to date there are no reports of the observation of any dielectric anomaly. The present study suggests that incorporation of the NH4þ ion ought to induce multiferroic character in these exotic magnetic lattices.

’ ASSOCIATED CONTENT

bS

Supporting Information. Crystallographic data (CIF). This material is available free of charge via the Internet at http:// pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

[email protected] 3794

dx.doi.org/10.1021/ja1117683 |J. Am. Chem. Soc. 2011, 133, 3792–3795

Journal of the American Chemical Society

COMMUNICATION

’ ACKNOWLEDGMENT This work was supported by NSF-DMR Grant 0506946. We thank the National High Magnetic Field Laboratory, Tallahassee, for use of the experimental facilities. NHMFL is supported by NSF cooperative agreement No. DMR-0654118, by the State of Florida, and the DOE. ’ REFERENCES (1) (a) Spaldin, N. A.; Cheong, S.; Ramesh, R. Phys. Today 2010, 63 (10), 38. (b) Cheetham, A. K.; Rao, C. N. R. Science 2007, 318, 58. (2) (a) Jain, P.; Dalal, N. S.; Toby, B. H.; Kroto, H. W.; Cheetham, A. K. J. Am. Chem. Soc. 2008, 130, 10450. (b) Jain, P.; Ramachandran, V.; Clark, R. J.; Zhou, H. D.; Toby, B. H.; Dalal, N. S.; Kroto, H. W.; Cheetham, A. K. J. Am. Chem. Soc. 2009, 131, 13625. (3) Riesenfeld, E. H. Chem. Ber. 1905, 38, 4068. (4) Dickman, M. H.; Pope, M. T. Chem. Rev. 1994, 94, 569. (5) (a) Dalal, N. S.; Millar, J. M.; Jagadeesh, M. S.; Seehra, M. S. J. Chem. Phys. 1981, 74, 1916. (b) Cage, B.; Dalal, N. S. Chem. Mater. 2001, 13, 880. (6) (a) Choi, K.-Y.; Nellutla, S.; Reyes, A. P.; Kuhns, P. L.; Jo, Y.-J.; Balicas, L.; Nojiri, H.; Pati, M.; Dalal, N. S. Phys. Rev. B. 2008, 78, No. 214419. (b) Nellutla, S.; Pati, M.; Jo, Y.-J.; Zhou, H. D.; Moon, B. H.; Pajerowski, D. M.; Yoshida, Y.; Janik, J. A.; Balicas, L.; Lee, Y.; Miesel, M. W.; Takano, Y.; Wiebe, C. R.; Dalal, N. S. Phys. Rev. B 2010, 81, No. 064431. (7) Harter, A. G.; Cage, B.; Nguyen, P.; Abboud, K. A.; Dalal, N. S. Polyhedron 2005, 24, 2350. (8) Lines, M. E.; Glass, A. M. Principles and Applications of Ferroelectrics and Related Materials; Oxford University Press; New York, 1977. (9) (a) Lasave, J.; Koval, S.; Dalal, N. S.; Migoni, R. L. Phys. Rev. Lett. 2007, 98, No. 267601. (b) Dalal, N.; Klymachyov, A.; BussmannHolder, A. Phys. Rev. Lett. 1998, 81, 5924. (10) (a) Shiramura, W.; Takatsu, K.; Kurniawan, B.; Tanaka, H.; Uekusa, H.; Ohashi, Y.; Takizawa, K.; Mitamura, H.; Goto, T. J. Phys. Soc. Jpn. 1998, 67, 1548. (b) Ruegg, C.; Oettli, M.; Schefer, J.; Zaharko, O.; Furrer, A.; Tanaka, H.; Kr€amer, K. W.; G€udel, H. U.; Vorderwisch, P.; Habicht, K.; Polinski, T.; Meissner, M. Phys. Rev. Lett. 2004, 93, No. 037207. (c) Hosoya, S.; Inokuchi, K.; Suzuki, T.; Goto, T.; Tanaka, H.; Awaji, S.; Watanabe, K.; Sasaki, T.; Fukase, T.; Shimizu, T.; Goto, A. Physica B 2003, 329, 977.

3795

dx.doi.org/10.1021/ja1117683 |J. Am. Chem. Soc. 2011, 133, 3792–3795